OLD | NEW |
(Empty) | |
| 1 from __future__ import unicode_literals |
| 2 from __future__ import absolute_import |
| 3 from . import util |
| 4 from . import odict |
| 5 |
| 6 |
| 7 class State(list): |
| 8 """ Track the current and nested state of the parser. |
| 9 |
| 10 This utility class is used to track the state of the BlockParser and |
| 11 support multiple levels if nesting. It's just a simple API wrapped around |
| 12 a list. Each time a state is set, that state is appended to the end of the |
| 13 list. Each time a state is reset, that state is removed from the end of |
| 14 the list. |
| 15 |
| 16 Therefore, each time a state is set for a nested block, that state must be |
| 17 reset when we back out of that level of nesting or the state could be |
| 18 corrupted. |
| 19 |
| 20 While all the methods of a list object are available, only the three |
| 21 defined below need be used. |
| 22 |
| 23 """ |
| 24 |
| 25 def set(self, state): |
| 26 """ Set a new state. """ |
| 27 self.append(state) |
| 28 |
| 29 def reset(self): |
| 30 """ Step back one step in nested state. """ |
| 31 self.pop() |
| 32 |
| 33 def isstate(self, state): |
| 34 """ Test that top (current) level is of given state. """ |
| 35 if len(self): |
| 36 return self[-1] == state |
| 37 else: |
| 38 return False |
| 39 |
| 40 |
| 41 class BlockParser: |
| 42 """ Parse Markdown blocks into an ElementTree object. |
| 43 |
| 44 A wrapper class that stitches the various BlockProcessors together, |
| 45 looping through them and creating an ElementTree object. |
| 46 """ |
| 47 |
| 48 def __init__(self, markdown): |
| 49 self.blockprocessors = odict.OrderedDict() |
| 50 self.state = State() |
| 51 self.markdown = markdown |
| 52 |
| 53 def parseDocument(self, lines): |
| 54 """ Parse a markdown document into an ElementTree. |
| 55 |
| 56 Given a list of lines, an ElementTree object (not just a parent |
| 57 Element) is created and the root element is passed to the parser |
| 58 as the parent. The ElementTree object is returned. |
| 59 |
| 60 This should only be called on an entire document, not pieces. |
| 61 |
| 62 """ |
| 63 # Create a ElementTree from the lines |
| 64 self.root = util.etree.Element(self.markdown.doc_tag) |
| 65 self.parseChunk(self.root, '\n'.join(lines)) |
| 66 return util.etree.ElementTree(self.root) |
| 67 |
| 68 def parseChunk(self, parent, text): |
| 69 """ Parse a chunk of markdown text and attach to given etree node. |
| 70 |
| 71 While the ``text`` argument is generally assumed to contain multiple |
| 72 blocks which will be split on blank lines, it could contain only one |
| 73 block. Generally, this method would be called by extensions when |
| 74 block parsing is required. |
| 75 |
| 76 The ``parent`` etree Element passed in is altered in place. |
| 77 Nothing is returned. |
| 78 |
| 79 """ |
| 80 self.parseBlocks(parent, text.split('\n\n')) |
| 81 |
| 82 def parseBlocks(self, parent, blocks): |
| 83 """ Process blocks of markdown text and attach to given etree node. |
| 84 |
| 85 Given a list of ``blocks``, each blockprocessor is stepped through |
| 86 until there are no blocks left. While an extension could potentially |
| 87 call this method directly, it's generally expected to be used |
| 88 internally. |
| 89 |
| 90 This is a public method as an extension may need to add/alter |
| 91 additional BlockProcessors which call this method to recursively |
| 92 parse a nested block. |
| 93 |
| 94 """ |
| 95 while blocks: |
| 96 for processor in self.blockprocessors.values(): |
| 97 if processor.test(parent, blocks[0]): |
| 98 if processor.run(parent, blocks) is not False: |
| 99 # run returns True or None |
| 100 break |
OLD | NEW |